It's called container and it is released by Apple at https://github.com/apple/container for Apple Silicon Mac devices.

This week we explore BERTopic + stable claim IDs:

• Use pre-computed embeddings from BigQuery (same pipeline as before).
• Fit/Load a BERTopic model (UMAP + HDBSCAN) in Python.
• Assign internal cluster IDs per batch, then map them to stable claim_ids.
• Persist to video_claims, claim_registry, and dim_claims tables for analysis.
• Inspect behavior in Looker Studio and reflect on limitations.

This week we explore BERTopic + stable topic IDs (via an ID registry):

• Train a BERTopic model in Python (UMAP + HDBSCAN).
• Map BERTopic’s internal clusters (model_version, internal_topic_id)
• Ensure topic IDs remain consistent across retraining (no more ID jumps).
• Join human-readable labels and persist results into video_topics for analysis.
• Inspect results in Looker Studio and reflect on limitations.

This week we explore embeddings + clustering:

• Generate embeddings of canonical topics and claims with BigQuery ML.
• Train a KMeans model on those embeddings to group semantically similar entries.
• Assign clusters back to each topic/claim.
• Inspect first results in Looker Studio and reflect on limitations.

The focus is on methods and reproducibility, not on polished production code. The snippets below are meant as guidance scaffolding, but already allow you to build a similar pipeline.

There is e.g. payload-preferences or payload-migrations. Since 3.0 there is also payload-jobs for the neat queue system and payload-locked-documents for the document locking.

But if you want to develop a directus endpoint extension locally, you might want to use the "breakpoint" feature of your IDE (e.g. vscode). Without the need to run the entire directus stack with pnpm in development mode.

The official vscode extension for wsl usually integrates nicely with visual code.

But today I experienced a very bad performance of vscode and the windows system. A short htop / taskmanager call showed: All CPUs have been in use for 100%. The process eating all resource was called: .vscode-server.

So we figured that this did not work anymore in the newer version of dotenv (see commit in 4.3) the default value for use_putenv changed to false. And in symfony 5 there was a breaking change to switch use_putenv from true to false (see the putenv deprecation pull request.

However, when I pulled the latest tar.gz (v10.4.0.316 as of 2023/01/13) at https://github.com/newrelic/newrelic-php-agent/archive/refs/tags/v10.4.0.316.tar.gz and tried to compile - it did nearly work.

The first tries of make agent, will fail with missing pcre_compile or aclocal or glibtoolize. This can be fixed by installing brew install pcre for pcre_compile, brew install automake for aclocal and brew install libtool for glibtoolize.

Then finally running make agent ends up in:

util_hash.c:198:5: error: unannotated fall-through between switch labels [-Werror,-Wimplicit-fallthrough]
    case 2:
    ^
util_hash.c:198:5: note: insert '__attribute__((fallthrough));' to silence this warning
    case 2:
    ^
    __attribute__((fallthrough));
util_hash.c:198:5: note: insert 'break;' to avoid fall-through
    case 2:
    ^
    break;